# Generative AI Risks [[Generative AI (Gen AI)]] have associated benefits, but also risks to consider: - Confidentiality and security: sensitive data is at risk because LLMs may be trained on the data you feed them, meaning that it could be inadvertently exposed - Bias and discrimination: Generative AI models can inherit and even amplify biases present in their training data - Hallucinations: Generative AI may generate outputs that sound plausible, but factually incorrect or even nonsensical - Environmental: Generative AI models are power hungry. They have an important ecological impact - Social: Generative AI might disrupt traditional social structure and constructs - Reputation: Using Generative AI may harm the reputation of organizations and individuals - Liability: Using Generative AI may lead to increase liabilities for organizations and individuals. If AI is doing your job and you don't master the output, then you are at risk - Compliance: Using Generative AI requires ensuring compliancy with various regulations. Not everyone can do so without proper consideration It's of critical importance to keep humans in the loop when using Generative AI. Human oversight is essential to ensure the responsible and ethical use of these technologies.